Showing 1–16 of 16 results for author: Zhdanov, M

Electrostatics from Laplacian Eigenbasis for Neural Network Interatomic Potentials

Authors: Maksim Zhdanov, Vladislav Kurenkov

Abstract: Recent advances in neural network interatomic potentials have emerged as a promising research direction. However, popular deep learning models often lack auxiliary constraints grounded in physical laws, which could accelerate training and improve fidelity through physics-based regularization. In this work, we introduce $Φ$-Module, a universal plugin module that enforces Poisson's equation within t… ▽ More Recent advances in neural network interatomic potentials have emerged as a promising research direction. However, popular deep learning models often lack auxiliary constraints grounded in physical laws, which could accelerate training and improve fidelity through physics-based regularization. In this work, we introduce $Φ$-Module, a universal plugin module that enforces Poisson's equation within the message-passing framework to learn electrostatic interactions in a self-supervised manner. Specifically, each atom-wise representation is encouraged to satisfy a discretized Poisson's equation, making it possible to acquire a potential $\boldsymbolφ$ and a corresponding charge density $\boldsymbolρ$ linked to the learnable Laplacian eigenbasis coefficients of a given molecular graph. We then derive an electrostatic energy term, crucial for improved total energy predictions. This approach integrates seamlessly into any existing neural potential with insignificant computational overhead. Experiments on the OE62 and MD22 benchmarks confirm that models combined with $Φ$-Module achieve robust improvements over baseline counterparts. For OE62 error reduction ranges from 4.5\% to 17.8\%, and for MD22, baseline equipped with $Φ$-Module achieves best results on 5 out of 14 cases. Our results underscore how embedding a first-principles constraint in neural interatomic potentials can significantly improve performance while remaining hyperparameter-friendly, memory-efficient and lightweight in training. Code will be available at \href{https://github.com/dunnolab/phi-module}{dunnolab/phi-module}. △ Less

Submitted 20 May, 2025; originally announced May 2025.

AdS-GNN -- a Conformally Equivariant Graph Neural Network

Authors: Maksim Zhdanov, Nabil Iqbal, Erik Bekkers, Patrick Forré

Abstract: Conformal symmetries, i.e.\ coordinate transformations that preserve angles, play a key role in many fields, including physics, mathematics, computer vision and (geometric) machine learning. Here we build a neural network that is equivariant under general conformal transformations. To achieve this, we lift data from flat Euclidean space to Anti de Sitter (AdS) space. This allows us to exploit a kn… ▽ More Conformal symmetries, i.e.\ coordinate transformations that preserve angles, play a key role in many fields, including physics, mathematics, computer vision and (geometric) machine learning. Here we build a neural network that is equivariant under general conformal transformations. To achieve this, we lift data from flat Euclidean space to Anti de Sitter (AdS) space. This allows us to exploit a known correspondence between conformal transformations of flat space and isometric transformations on the AdS space. We then build upon the fact that such isometric transformations have been extensively studied on general geometries in the geometric deep learning literature. We employ message-passing layers conditioned on the proper distance, yielding a computationally efficient framework. We validate our model on tasks from computer vision and statistical physics, demonstrating strong performance, improved generalization capacities, and the ability to extract conformal data such as scaling dimensions from the trained network. △ Less

Submitted 19 May, 2025; originally announced May 2025.

Social Dynamics: Parameter Estimate

Authors: Michael Zhdanov

Abstract: The goal of this paper is to estimate some parameters of simple social harmonic oscillations modelling U.S. Presidential Approval and Macropartisanship. The harmonic oscillations are simplest solutions of equations of the deterministic social dynamics (Zhdanov M., 2024), relating social bodies positions in the Bayesian space of assessments to noncoercive driving forces acting on these social bodie… ▽ More The goal of this paper is to estimate some parameters of simple social harmonic oscillations modelling U.S. Presidential Approval and Macropartisanship. The harmonic oscillations are simplest solutions of equations of the deterministic social dynamics (Zhdanov M., 2024), relating social bodies positions in the Bayesian space of assessments to noncoercive driving forces acting on these social bodies. In some limits, such as individuals assessing each other, the space may be reduced to just sets of scores, ratings, or indexes used to measure social variables like rates of presidential approval or consumer sentiments. Thereby, relations are established between forces of social attraction and elasticity, considered in the presented paper, and assessments of social bodies including individuals, by other social bodies. The results are applied to estimate parameters of social bodies and their motions from survey data on Presidential Approval, U.S. Senator Approval, and Macropartisanship and provide one more perspective on some well-known issues like what moves macropartisanship and why presidential approval fluctuates. △ Less

Submitted 25 February, 2025; originally announced February 2025.

Comments: 27 pages, 5 figures

MSC Class: 91D10

Erwin: A Tree-based Hierarchical Transformer for Large-scale Physical Systems

Authors: Maksim Zhdanov, Max Welling, Jan-Willem van de Meent

Abstract: Large-scale physical systems defined on irregular grids pose significant scalability challenges for deep learning methods, especially in the presence of long-range interactions and multi-scale coupling. Traditional approaches that compute all pairwise interactions, such as attention, become computationally prohibitive as they scale quadratically with the number of nodes. We present Erwin, a hierar… ▽ More Large-scale physical systems defined on irregular grids pose significant scalability challenges for deep learning methods, especially in the presence of long-range interactions and multi-scale coupling. Traditional approaches that compute all pairwise interactions, such as attention, become computationally prohibitive as they scale quadratically with the number of nodes. We present Erwin, a hierarchical transformer inspired by methods from computational many-body physics, which combines the efficiency of tree-based algorithms with the expressivity of attention mechanisms. Erwin employs ball tree partitioning to organize computation, which enables linear-time attention by processing nodes in parallel within local neighborhoods of fixed size. Through progressive coarsening and refinement of the ball tree structure, complemented by a novel cross-ball interaction mechanism, it captures both fine-grained local details and global features. We demonstrate Erwin's effectiveness across multiple domains, including cosmology, molecular dynamics, PDE solving, and particle fluid dynamics, where it consistently outperforms baseline methods both in accuracy and computational efficiency. △ Less

Submitted 2 June, 2025; v1 submitted 24 February, 2025; originally announced February 2025.

Comments: Accepted to ICML 2025. Code: https://github.com/maxxxzdn/erwin

Idealized Social Dynamics In Bayesian Space of Assessments

Authors: Michael Zhdanov

Abstract: The purpose of this paper is to present an idealized hypotheses-drawn model describing dynamics of variability in business and other social areas. A new construct is introduced called the Bayesian space of assessments to consider changes in positions of individuals and intrinsically interrelated entities called social bodies. The bodies spatial coordinates are their assessments in any required for… ▽ More The purpose of this paper is to present an idealized hypotheses-drawn model describing dynamics of variability in business and other social areas. A new construct is introduced called the Bayesian space of assessments to consider changes in positions of individuals and intrinsically interrelated entities called social bodies. The bodies spatial coordinates are their assessments in any required for a specific purpose professional and or ethical dimensions. A concept of market power introduced originally in entrepreneurship and technology commercialization (Michael Danov, Brock Smith, and Ronald Mitchell, 2003) is extended to become a driving force applicable to other social processes. The model describes interrelations between driving forces acting on social bodies and changes in positions of these bodies in the Bayesian space of assessments. Implications are discussed for some business and political oscillations. △ Less

Submitted 12 June, 2024; originally announced June 2024.

Comments: 46 pages, 2 figures

MSC Class: 91D10

Clifford-Steerable Convolutional Neural Networks

Authors: Maksim Zhdanov, David Ruhe, Maurice Weiler, Ana Lucic, Johannes Brandstetter, Patrick Forré

Abstract: We present Clifford-Steerable Convolutional Neural Networks (CS-CNNs), a novel class of $\mathrm{E}(p, q)$-equivariant CNNs. CS-CNNs process multivector fields on pseudo-Euclidean spaces $\mathbb{R}^{p,q}$. They cover, for instance, $\mathrm{E}(3)$-equivariance on $\mathbb{R}^3$ and Poincaré-equivariance on Minkowski spacetime $\mathbb{R}^{1,3}$. Our approach is based on an implicit parametrizatio… ▽ More We present Clifford-Steerable Convolutional Neural Networks (CS-CNNs), a novel class of $\mathrm{E}(p, q)$-equivariant CNNs. CS-CNNs process multivector fields on pseudo-Euclidean spaces $\mathbb{R}^{p,q}$. They cover, for instance, $\mathrm{E}(3)$-equivariance on $\mathbb{R}^3$ and Poincaré-equivariance on Minkowski spacetime $\mathbb{R}^{1,3}$. Our approach is based on an implicit parametrization of $\mathrm{O}(p,q)$-steerable kernels via Clifford group equivariant neural networks. We significantly and consistently outperform baseline methods on fluid dynamics as well as relativistic electrodynamics forecasting tasks. △ Less

Submitted 6 July, 2024; v1 submitted 22 February, 2024; originally announced February 2024.

Comments: accepted to ICML 2024

Identity Curvature Laplace Approximation for Improved Out-of-Distribution Detection

Authors: Maksim Zhdanov, Stanislav Dereka, Sergey Kolesnikov

Abstract: Uncertainty estimation is crucial in safety-critical applications, where robust out-of-distribution (OOD) detection is essential. Traditional Bayesian methods, though effective, are often hindered by high computational demands. As an alternative, Laplace approximation offers a more practical and efficient approach to uncertainty estimation. In this paper, we introduce the Identity Curvature Laplac… ▽ More Uncertainty estimation is crucial in safety-critical applications, where robust out-of-distribution (OOD) detection is essential. Traditional Bayesian methods, though effective, are often hindered by high computational demands. As an alternative, Laplace approximation offers a more practical and efficient approach to uncertainty estimation. In this paper, we introduce the Identity Curvature Laplace Approximation (ICLA), a novel method that challenges the conventional posterior covariance formulation by using identity curvature and optimizing prior precision. This innovative design significantly enhances OOD detection performance on well-known datasets such as CIFAR-10, CIFAR-100, and ImageNet, while maintaining calibration scores. We attribute this improvement to the alignment issues between typical feature embeddings and curvature as measured by the Fisher information matrix. Our findings are further supported by demonstrating that incorporating Fisher penalty or sharpness-aware minimization techniques can greatly enhance the uncertainty estimation capabilities of standard Laplace approximation. △ Less

Submitted 5 November, 2024; v1 submitted 16 December, 2023; originally announced December 2023.

Catching Image Retrieval Generalization

Authors: Maksim Zhdanov, Ivan Karpukhin

Abstract: The concepts of overfitting and generalization are vital for evaluating machine learning models. In this work, we show that the popular Recall@K metric depends on the number of classes in the dataset, which limits its ability to estimate generalization. To fix this issue, we propose a new metric, which measures retrieval performance, and, unlike Recall@K, estimates generalization. We apply the pro… ▽ More The concepts of overfitting and generalization are vital for evaluating machine learning models. In this work, we show that the popular Recall@K metric depends on the number of classes in the dataset, which limits its ability to estimate generalization. To fix this issue, we propose a new metric, which measures retrieval performance, and, unlike Recall@K, estimates generalization. We apply the proposed metric to popular image retrieval methods and provide new insights about deep metric learning generalization. △ Less

Submitted 23 June, 2023; originally announced June 2023.

Comments: 4 pages, 3 figures, 2 tables

Machine learning-assisted close-set X-ray diffraction phase identification of transition metals

Authors: Maksim Zhdanov, Andrey Zhdanov

Abstract: Machine learning has been applied to the problem of X-ray diffraction phase prediction with promising results. In this paper, we describe a method for using machine learning to predict crystal structure phases from X-ray diffraction data of transition metals and their oxides. We evaluate the performance of our method and compare the variety of its settings. Our results demonstrate that the propose… ▽ More Machine learning has been applied to the problem of X-ray diffraction phase prediction with promising results. In this paper, we describe a method for using machine learning to predict crystal structure phases from X-ray diffraction data of transition metals and their oxides. We evaluate the performance of our method and compare the variety of its settings. Our results demonstrate that the proposed machine learning framework achieves competitive performance. This demonstrates the potential for machine learning to significantly impact the field of X-ray diffraction and crystal structure determination. Open-source implementation: https://github.com/maxnygma/NeuralXRD. △ Less

Submitted 28 April, 2023; originally announced May 2023.

Comments: Published at the ICLR 2023 Workshop on Machine Learning for Materials, 5 pages, 1 figure

Diversifying Deep Ensembles: A Saliency Map Approach for Enhanced OOD Detection, Calibration, and Accuracy

Authors: Stanislav Dereka, Ivan Karpukhin, Maksim Zhdanov, Sergey Kolesnikov

Abstract: Deep ensembles are capable of achieving state-of-the-art results in classification and out-of-distribution (OOD) detection. However, their effectiveness is limited due to the homogeneity of learned patterns within ensembles. To overcome this issue, our study introduces Saliency Diversified Deep Ensemble (SDDE), a novel approach that promotes diversity among ensemble members by leveraging saliency… ▽ More Deep ensembles are capable of achieving state-of-the-art results in classification and out-of-distribution (OOD) detection. However, their effectiveness is limited due to the homogeneity of learned patterns within ensembles. To overcome this issue, our study introduces Saliency Diversified Deep Ensemble (SDDE), a novel approach that promotes diversity among ensemble members by leveraging saliency maps. Through incorporating saliency map diversification, our method outperforms conventional ensemble techniques and improves calibration in multiple classification and OOD detection tasks. In particular, the proposed method achieves state-of-the-art OOD detection quality, calibration, and accuracy on multiple benchmarks, including CIFAR10/100 and large-scale ImageNet datasets. △ Less

Submitted 5 November, 2024; v1 submitted 19 May, 2023; originally announced May 2023.

Implicit Convolutional Kernels for Steerable CNNs

Authors: Maksim Zhdanov, Nico Hoffmann, Gabriele Cesa

Abstract: Steerable convolutional neural networks (CNNs) provide a general framework for building neural networks equivariant to translations and transformations of an origin-preserving group $G$, such as reflections and rotations. They rely on standard convolutions with $G$-steerable kernels obtained by analytically solving the group-specific equivariance constraint imposed onto the kernel space. As the so… ▽ More Steerable convolutional neural networks (CNNs) provide a general framework for building neural networks equivariant to translations and transformations of an origin-preserving group $G$, such as reflections and rotations. They rely on standard convolutions with $G$-steerable kernels obtained by analytically solving the group-specific equivariance constraint imposed onto the kernel space. As the solution is tailored to a particular group $G$, implementing a kernel basis does not generalize to other symmetry transformations, complicating the development of general group equivariant models. We propose using implicit neural representation via multi-layer perceptrons (MLPs) to parameterize $G$-steerable kernels. The resulting framework offers a simple and flexible way to implement Steerable CNNs and generalizes to any group $G$ for which a $G$-equivariant MLP can be built. We prove the effectiveness of our method on multiple tasks, including N-body simulations, point cloud classification and molecular property prediction. △ Less

Submitted 27 October, 2023; v1 submitted 12 December, 2022; originally announced December 2022.

Comments: Accepted to 37th Conference on Neural Information Processing Systems (NeurIPS 2023)

Amortized Bayesian Inference of GISAXS Data with Normalizing Flows

Authors: Maksim Zhdanov, Lisa Randolph, Thomas Kluge, Motoaki Nakatsutsumi, Christian Gutt, Marina Ganeva, Nico Hoffmann

Abstract: Grazing-Incidence Small-Angle X-ray Scattering (GISAXS) is a modern imaging technique used in material research to study nanoscale materials. Reconstruction of the parameters of an imaged object imposes an ill-posed inverse problem that is further complicated when only an in-plane GISAXS signal is available. Traditionally used inference algorithms such as Approximate Bayesian Computation (ABC) rel… ▽ More Grazing-Incidence Small-Angle X-ray Scattering (GISAXS) is a modern imaging technique used in material research to study nanoscale materials. Reconstruction of the parameters of an imaged object imposes an ill-posed inverse problem that is further complicated when only an in-plane GISAXS signal is available. Traditionally used inference algorithms such as Approximate Bayesian Computation (ABC) rely on computationally expensive scattering simulation software, rendering analysis highly time-consuming. We propose a simulation-based framework that combines variational auto-encoders and normalizing flows to estimate the posterior distribution of object parameters given its GISAXS data. We apply the inference pipeline to experimental data and demonstrate that our method reduces the inference cost by orders of magnitude while producing consistent results with ABC. △ Less

Submitted 4 October, 2022; originally announced October 2022.

Learning Generative Factors of EEG Data with Variational auto-encoders

Authors: Maksim Zhdanov, Saskia Steinmann, Nico Hoffmann

Abstract: Electroencephalography produces high-dimensional, stochastic data from which it might be challenging to extract high-level knowledge about the phenomena of interest. We address this challenge by applying the framework of variational auto-encoders to 1) classify multiple pathologies and 2) recover the neurological mechanisms of those pathologies in a data-driven manner. Our framework learns generat… ▽ More Electroencephalography produces high-dimensional, stochastic data from which it might be challenging to extract high-level knowledge about the phenomena of interest. We address this challenge by applying the framework of variational auto-encoders to 1) classify multiple pathologies and 2) recover the neurological mechanisms of those pathologies in a data-driven manner. Our framework learns generative factors of data related to pathologies. We provide an algorithm to decode those factors further and discover how different pathologies affect observed data. We illustrate the applicability of the proposed approach to identifying schizophrenia, either followed or not by auditory verbal hallucinations. We further demonstrate the ability of the framework to learn disease-related mechanisms consistent with current domain knowledge. We also compare the proposed framework with several benchmark approaches and indicate its classification performance and interpretability advantages. △ Less

Submitted 17 August, 2022; v1 submitted 4 June, 2022; originally announced June 2022.

Comments: Accepted to DGM4MICCAI workshop at MICCAI 2022

Investigating Brain Connectivity with Graph Neural Networks and GNNExplainer

Authors: Maksim Zhdanov, Saskia Steinmann, Nico Hoffmann

Abstract: Functional connectivity plays an essential role in modern neuroscience. The modality sheds light on the brain's functional and structural aspects, including mechanisms behind multiple pathologies. One such pathology is schizophrenia which is often followed by auditory verbal hallucinations. The latter is commonly studied by observing functional connectivity during speech processing. In this work,… ▽ More Functional connectivity plays an essential role in modern neuroscience. The modality sheds light on the brain's functional and structural aspects, including mechanisms behind multiple pathologies. One such pathology is schizophrenia which is often followed by auditory verbal hallucinations. The latter is commonly studied by observing functional connectivity during speech processing. In this work, we have made a step toward an in-depth examination of functional connectivity during a dichotic listening task via deep learning for three groups of people: schizophrenia patients with and without auditory verbal hallucinations and healthy controls. We propose a graph neural network-based framework within which we represent EEG data as signals in the graph domain. The framework allows one to 1) predict a brain mental disorder based on EEG recording, 2) differentiate the listening state from the resting state for each group and 3) recognize characteristic task-depending connectivity. Experimental results show that the proposed model can differentiate between the above groups with state-of-the-art performance. Besides, it provides a researcher with meaningful information regarding each group's functional connectivity, which we validated on the current domain knowledge. △ Less

Submitted 4 June, 2022; originally announced June 2022.

Comments: Submitted to ICPR 2022

Impact of the improved parallel kinetic coefficients on the helium and neon transport in SOLPS-ITER for ITER

Authors: S. O. Makarov, D. P. Coster, V. A. Rozhansky, S. P. Voskoboynikov, E. G. Kaveeva, I. Y. Senichenkov, A. A. Stepanenko, V. M. Zhdanov, X. Bonnin

Abstract: New Grad's-Zhdanov module is implemented in the SOLPS-ITER code and applied to ITER impurity transport simulations. Significant difference appears in the helium transport due to improved parallel kinetic coefficients. As a result 30\% decrease of the separatrix-averaged helium relative concentration is observed for the constant helium source and pumping speed. Change of the impurity behaviour is d… ▽ More New Grad's-Zhdanov module is implemented in the SOLPS-ITER code and applied to ITER impurity transport simulations. Significant difference appears in the helium transport due to improved parallel kinetic coefficients. As a result 30\% decrease of the separatrix-averaged helium relative concentration is observed for the constant helium source and pumping speed. Change of the impurity behaviour is discussed. For the neon changes are less pronounced. For the first time the ion distribution functions are studied in the ITER Scrape-off layer conditions to reveal the origin of the kinetic coefficient improvements and theory limitations. △ Less

Submitted 14 January, 2022; originally announced January 2022.

Comments: This is the pre-submission version. The published version can be found using DOI

Journal ref: Contributions to Plasma Physics 2021

Detection of nanocracks on double fluoride rare earth crystal surface

Authors: R. Yu. Abdulsabirov, A. A. Bukharaev, M. R. Zhdanov, R. Sh. Zhdanov, A. V. Klochkov, S. L. Korableva, V. V. Naletov, N. I. Nurgazizov, M. S. Tagirov, D. A. Tayurskii

Abstract: Predicted earlier, microcracks on the crystal surface of both finely dispersed $LiYF_4$ powders and single crystals of the Van Vleck paramagnet $% LiTmF_4$ were detected by using the NMR Cryoporometry and Atomic-Force Microscopy technique. Predicted earlier, microcracks on the crystal surface of both finely dispersed $LiYF_4$ powders and single crystals of the Van Vleck paramagnet $% LiTmF_4$ were detected by using the NMR Cryoporometry and Atomic-Force Microscopy technique. △ Less

Submitted 15 August, 1998; originally announced August 1998.

Comments: 5 pages, 2 Postscript and 3 JPEG figures, LaTEX

Journal ref: Applied Magnetic Resonance, Volume 19, Number 2, Pages 197-208, 2000